It is known that the occurrence of propagation loss, or “pathloss”, due to the scattering or absorption of a radio communication as it travels through space, causes the strength of a signal to diminish. Factors which influence the pathloss between a transmitter and a receiver include: transmitter antenna height, receiver antenna height, carrier frequency, clutter type (urban, sub-urban, rural), details of morphology such as height, density, separation, terrain type (hilly, flat). The pathloss L (dB) between a transmitter and a receiver can be modelled by:L=b+10n log d  (A)
Where d (meters) is the transmitter-receiver separation, b(db) and n are the pathloss parameters and the absolute pathloss is given by l=10(L/10) 
FIG. 1A illustrates a single-cell two-hop wireless communication system comprising a base station (known in the context of 3 G communication systems as “node-B” (NB)) a relay node (RN) and a user equipment (UE). In the case where signals are being transmitted on the downlink (DL) from a base station to a destination user equipment (UE) via the relay node (RN), the base station comprises the source apparatus (S) and the user equipment comprises the destination apparatus (D). In the case where communication signals are being transmitted on the uplink (UL) from user equipment (UE), via the relay node, to the base station, the user equipment comprises the source apparatus and the base station comprises the destination apparatus. The relay node is an example of an intermediate apparatus (I) and comprises: a receiver, operable to receive a signal from the source apparatus; and a transmitter, operable to transmit this signal, or a derivative thereof, to the destination apparatus.
Table I below gives some examples of the calculated pathloss of a signal being transmitted over the different links: source to destination (SD), source to intermediate
(SI) and intermediate to destination (ID), in a multi-hop transmission system where b and n are assumed to remain the same over each of the links.
TABLE ISeparation(metres)Pathloss in dBAbsolute Pathlossb(dB)nSDSIIDSDSIIDSDSIID15.33.761000500500128.1116.8116.86.46E124.77E114.77E1115.33.761000600600128.1119.76119.766.46E129.46E119.46E1115.33.761000700700128.1122.28122.286.46E121.69E121.69E12
The examples calculated above demonstrate that the sum of the absolute path losses experienced over the indirect link SI+ID may be less than the pathloss experienced over the direct link SD. In other words it is possible for:L(SI)+L(ID)<L(SD)  (B)
Splitting a single transmission link into two shorter transmission segments therefore exploits the non-linear relationship between pathloss verses distance. From a simple theoretical analysis of the pathloss using equation (A), it can be appreciated that a reduction in the overall pathloss (and therefore an improvement, or gain, in signal strength and thus data throughput) should be achieved if a signal is sent from a source apparatus to a destination apparatus via an intermediate apparatus (eg relay node), rather than being sent directly from the source apparatus to the destination apparatus. If implemented, multi-hop communication systems could potentially allow for a reduction in the transmit power of transmitters which facilitate wireless transmissions, which would lead to a reduction in interference levels as well as decreasing exposure to electromagnetic emissions.
Clearly, due to the non-linear relationship between pathloss and distance, the position of an intermediate apparatus relative to the source and destination, will critically effect the potential gain that a multi-hop transmission may have as compared to a direct, or single-hop, transmission between the source and destination. This is illustrated in FIG. 2A which shows a graphical representation of the theoretical gain which may be achieved by multi-hop transmissions, and plots the total power loss (dB) against the relative normalized position of the intermediate apparatus between the source apparatus and the destination apparatus.
Considering firstly the case where the intermediate node is positioned on the line of the direct link between the source and destination (in which case the path extension factor (s)=1), it can be seen that the potential gain is reduced as the relay node is moved away from a mid-way position towards the source or destination apparatus. Likewise, as the position of the intermediate apparatus is moved away from the line of the direct link, thereby extending the total path length of the sum of the two transmission segments (and increasing the path extension factor to s=1.1, s=1.2 etc), it can be seen that the graphical region of theoretical gain is again reduced.
However, simulations carried out to test the applicability of multi-hop communication systems have revealed unexpectedly low gains in throughput of data. Indeed, the gains experienced are well below the potential gain suggested by a simple analysis based on the pathloss equation A. Consequently, and despite the potential advantages that multi-hop systems may demonstrate in terms of signal range extension, a possible reduction in the overall transmit power required to transmit a signal between a source and destination, and the connectivity of otherwise inaccessible nodes, wireless systems operators have been deterred from implementing multi-hop networks.
One of the reasons that such a discrepancy exists between the predicted gain and the simulated gain is that previous predictions have been based on the assumption that the pathloss parameters b and n are the same on all links. In actual fact, these values vary as a result of the antenna height of the source apparatus and destination apparatus as compared to the height of the relay node. Thus, a more realistic table of values is given below in table II. The values labelled 3GPP are obtained from adapting the model employed by the 3GPP to incorporate the fact that the antenna height of the intermediate apparatus is typically somewhere between the height of the antenna at the source and destination apparatus. The values labelled UoB are derived from modelling conducted by the University of Bristol based on a typical deployment in the city of Bristol.
TABLE IILinkPathloss ParameterS-DS-II-D3GPPb (dB)15.315.528n3.763.684UoBb (dB)13.0716.2910.04n4.884.645.47
The graphical illustration of total pathloss verses normalized relay node position using the pathloss parameters tabulated in table II is shown in FIG. 2B. It can be seen that the perfect “bell-shape” of FIG. 2A is not achieved when a more realistic set of pathloss parameters are used to calculate the variation in total pathloss as the position of a theoretical relay node is adjusted. Indeed, the region of gain is reduced and it is apparent that relatively small changes in the position of a relay node or a user equipment, leading to a change in the absolute pathloss over the communication link, will have a significant effect on the quality of a communication signal at the receiving apparatus. Thus, the positioning of an intermediate apparatus or relay node is critical if a gain is to be achieved by the occurrence of a multi-hop transmission, as compared to a direct transmission between the source and destination.
However, even when predictions are based on a more accurate reflection of the pathloss parameters likely to be encountered in the real world, simulations of multi-hop systems have revealed unexpectedly poor correspondence between the predicted and simulated gain.